Adapt or Die: AI’s Reckoning for Quant Funds

The Darwinism of Data

Feng Ji, founder of Baiont and former WorldQuant executive, delivers a stark message to the financial industry: Quantitative fund managers who do not embrace AI and machine learning will be “eliminated by the market.” In a conversation with the Financial Times, Ji outlines how advanced AI tools—especially large language models—are turbocharging signal generation and drastically reducing inefficiencies in global markets. As traditional quant signals become commoditized and margins narrow, those who fail to innovate using AI will struggle to stay competitive. Ji argues that markets have become more efficient and zero-sum, requiring managers to “move faster” and embrace disruptive tools to remain relevant.

New Tools, New Edge

Baiont, Ji’s AI-focused hedge fund launched last year, is designed with AI at its core. The firm leverages synthetic data and language models to generate high-frequency, high-quality investment signals across global equities. Ji contends that while traditional quants still rely on curated historical datasets, AI enables a more dynamic and nuanced understanding of unstructured data, such as text and sentiment. This technological edge, he believes, is not just a temporary advantage—it’s the future. Like code in computing, data in finance must now be processed at scale and in real time, and managers must build infrastructure that can adapt immediately as conditions change.

The Talent Revolution

Building such capabilities requires a shift not only in infrastructure but in mindset and recruitment. Ji likens today’s investor role to an AI “prompt engineer,” suggesting that future quant managers must be deeply conversant in both financial logic and machine learning techniques.